
import numpy as np
import pandas as pd
import datetime as dt
from pandas_datareader import data as wb
import plotly.graph_objects as go
import matplotlib.pyplot as plt
import seaborn as sn
import yfinance as yf
%matplotlib inline


stocks = input().split()
asset = [stock.upper() + '.NS' for stock in stocks]
asset

#Example: SBIN HDFC  ONGC WIPRO HEROMOTOCO LT ITC COALINDIA RELIANCE NCC

SBIN HDFC  ONGC WIPRO HEROMOTOCO LT ITC COALINDIA RELIANCE NCC

['SBIN.NS',
 'HDFC.NS',
 'ONGC.NS',
 'WIPRO.NS',
 'HEROMOTOCO.NS',
 'LT.NS',
 'ITC.NS',
 'COALINDIA.NS',
 'RELIANCE.NS',
 'NCC.NS']


asset.sort()


years = float(input('No. of year data you want to analyze: '))

## Example: 3 or 5 or 8 or 10

No. of year data you want to analyze: 5


endDate = dt.datetime.now()
startDate = endDate - dt.timedelta(days= 365 * years)
startDate

datetime.datetime(2018, 3, 31, 2, 2, 41, 607026)


yf.pdr_override()


pf_data = wb.get_data_yahoo(asset, start=startDate, end=endDate)['Adj Close']

[*********************100%***********************]  10 of 10 completed


pf_data


(pf_data / pf_data.iloc[0] * 100).plot(figsize= (10,5))
plt.show()


# Importing Required Library

from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential
from keras.layers import Dense, LSTM


## Scaling the data set as required

def Scale_data_set(data):

  dataset = data.values
  # Get the number of rows to train the model on
  training_data_len = int(np.ceil( len(dataset) * .95 ))

  scaler = MinMaxScaler(feature_range=(0,1))
  scaled_data = scaler.fit_transform(dataset)
  return training_data_len, scaled_data, dataset, scaler


## Creating a dataSet to be trained by LSTM

def Creating_training_data(training_data_len, scaled_data):
  # Create the training data set 
  # Create the scaled training data set
  train_data = scaled_data[0:int(training_data_len), :]
  # Split the data into x_train and y_train data sets
  x_train = []
  y_train = []

  for i in range(60, len(train_data)):
      x_train.append(train_data[i-60:i, 0])
      y_train.append(train_data[i, 0])
      
          
  # Convert the x_train and y_train to numpy arrays 
  x_train, y_train = np.array(x_train), np.array(y_train)

  # Reshape the data
  x_train = np.reshape(x_train, (x_train.shape[0], x_train.shape[1], 1))
  # x_train.shape

  return x_train, y_train


## Training by  LSTM

def LSTM_model(x_train, y_train):
  
  # Build the LSTM model
  model = Sequential()
  model.add(LSTM(128, return_sequences=True, input_shape= (x_train.shape[1], 1)))
  model.add(LSTM(64, return_sequences=False))
  model.add(Dense(25))
  model.add(Dense(1))

  # Compile the model
  model.compile(optimizer='adam', loss='mean_squared_error')

  # Train the model
  model.fit(x_train, y_train, batch_size=1, epochs=1)

  return model


## Predicting data using the trained value

def miscellaneous(model, training_data_len, scaled_data, dataset, scaler):
  # Create the testing data set
  # Create a new array containing scaled values from index 1543 to 2002 
  test_data = scaled_data[training_data_len - 60: , :]
  # Create the data sets x_test and y_test
  x_test = []
  y_test = dataset[training_data_len:, :]
  for i in range(60, len(test_data)):
      x_test.append(test_data[i-60:i, 0])
      
  # Convert the data to a numpy array
  x_test = np.array(x_test)

  # Reshape the data
  x_test = np.reshape(x_test, (x_test.shape[0], x_test.shape[1], 1 ))

  # Get the models predicted price values 
  predictions = model.predict(x_test)
  predictions = scaler.inverse_transform(predictions)

  # Get the root mean squared error (RMSE)
  rmse = np.sqrt(np.mean(((predictions - y_test) ** 2)))

  return predictions, rmse


## Ploting the chart with predicted values

def Ploting_Prediction(predictions ,training_data_len, data, title):
  # Plot the data
  train = data[:training_data_len]
  valid = data[training_data_len:]
  valid['Predictions'] = predictions
  # Visualize the data
  plt.figure(figsize=(16,6))
  plt.title(title)
  plt.xlabel('Date', fontsize=18)
  plt.ylabel('Close Price in Rs', fontsize=18)
  plt.plot(train['Close'])
  plt.plot(valid[['Close', 'Predictions']])
  plt.legend(['Train', 'Val', 'Predictions'], loc='lower right')
  plt.show()

  return train, valid

def plot_graph_plotly(predictions ,training_data_len, data, title):
  # Plot the data
  train = data[:training_data_len]
  valid = data[training_data_len:]
  valid['Predictions'] = predictions

  #Max Sharpe Ratio
  train_data = go.Scatter(
      name='Training data',
      
      x=data.index,
      y=train['Close'],
      marker=dict(color='blue',size=5,)
  )

  #Min Vol
  Val_data = go.Scatter(
      name='Actual Value',
     
      x=valid.index,
      y=valid['Close'],
      marker=dict(color='green',size=5,)
  )

  #Random portfolio
  Prediction_data = go.Scatter(
      name='Predicted Data',
      
      x=valid.index,
      y=valid['Predictions'],
      marker=dict(color='red',size=5,)
  )

  data = [train_data, Val_data, Prediction_data]

  layout = go.Layout(
      title = title,
      yaxis = dict(title='Close Price in Rs'),
      xaxis = dict(title='Date'),
      showlegend = True,
      legend = dict(
          x = .83, y = 0, traceorder='normal',
          bgcolor='#E2E2E2',
          bordercolor='black',
          borderwidth=2),
      width=980,
      height=500)

  fig = go.Figure(data=data, layout=layout)
  fig.show()

  return train, valid


##  Compiling all functions to one functions

def Stock_prediction(data, title):
  training_data_len, scaled_data, dataset, scaler = Scale_data_set(data)
  x_train, y_train = Creating_training_data(training_data_len, scaled_data)
  model = LSTM_model(x_train, y_train)
  predictions, rmse = miscellaneous(model, training_data_len, scaled_data, dataset, scaler)
  train, valid = plot_graph_plotly(predictions ,training_data_len, data, title)

  return train, valid, rmse, predictions


data = {}
for st in asset:
  data['{}'.format(st)] = pf_data.filter([st])


pf_data


for i in asset:
  DATA = data[i]
  DATA = DATA.rename(columns = {i:"Close"})
  train, valid, rmse, predictions = Stock_prediction(DATA, i)
  print("'root mean squared error' of {} = ".format(i) + str(rmse))

1116/1116 [==============================] - 31s 25ms/step - loss: 0.0023
2/2 [==============================] - 1s 33ms/step

C:\Users\DC\AppData\Local\Temp/ipykernel_8728/3901712836.py:24: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  valid['Predictions'] = predictions

'root mean squared error' of COALINDIA.NS = 10.86642588586606
1116/1116 [==============================] - 32s 26ms/step - loss: 0.0038
2/2 [==============================] - 1s 31ms/step

C:\Users\DC\AppData\Local\Temp/ipykernel_8728/3901712836.py:24: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

'root mean squared error' of HDFC.NS = 76.52052309278811
1116/1116 [==============================] - 35s 28ms/step - loss: 0.0048
2/2 [==============================] - 1s 38ms/step

C:\Users\DC\AppData\Local\Temp/ipykernel_8728/3901712836.py:24: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

'root mean squared error' of HEROMOTOCO.NS = 58.43621901908743
1116/1116 [==============================] - 34s 27ms/step - loss: 0.0015
2/2 [==============================] - 1s 42ms/step

C:\Users\DC\AppData\Local\Temp/ipykernel_8728/3901712836.py:24: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy

'root mean squared error' of ITC.NS = 21.208195699140397
1116/1116 [==============================] - 32s 25ms/step - loss: 0.0021
WARNING:tensorflow:5 out of the last 9 calls to <function Model.make_predict_function.<locals>.predict_function at 0x000001E736B903A0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has reduce_retracing=True option that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for  more details.
2/2 [==============================] - 1s 82ms/step

C:\Users\DC\AppData\Local\Temp/ipykernel_8728/3901712836.py:24: SettingWithCopyWarning:


A value is trying to be set on a copy of a slice from a DataFrame.
Try using .loc[row_indexer,col_indexer] = value instead

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy


retrn = (pf_data / pf_data.shift(1)) - 1
retrn


annual_returns = retrn.mean() * 250 *100
annual_returns.values[::-1].sort()
annual_returns

COALINDIA.NS     24.134615
HDFC.NS          21.721732
HEROMOTOCO.NS    15.547064
ITC.NS           15.474683
LT.NS            14.570013
NCC.NS           13.656731
ONGC.NS          13.356305
RELIANCE.NS       9.747047
SBIN.NS           8.729765
WIPRO.NS         -0.707604
dtype: float64


sorted_annual_return = annual_returns[0:5]
sorted_annual_return

COALINDIA.NS     24.134615
HDFC.NS          21.721732
HEROMOTOCO.NS    15.547064
ITC.NS           15.474683
LT.NS            14.570013
dtype: float64


assets = list(sorted_annual_return.index)
assets

['COALINDIA.NS', 'HDFC.NS', 'HEROMOTOCO.NS', 'ITC.NS', 'LT.NS']


pf_data = pf_data.filter(assets)
pf_data


returns = retrn.filter(assets)
returns


num_assets = len(assets)
num_assets

5


weights = np.random.random(num_assets)
weights /= np.sum(weights)
weights

array([0.2311764 , 0.24094695, 0.19052129, 0.1108554 , 0.22649996])


np.sum(weights * returns.mean()) * 250

0.10329953963472441


np.dot(weights.T, np.dot(returns.cov() * 250, weights))

0.045029652378750305


np.sqrt(np.dot(weights.T,np.dot(returns.cov() * 250, weights)))

0.21220191417315326


corr_matrix = returns.corr()
corr_matrix


pfolio_returns = []
pfolio_volatilities = []
pfolio_weights = []

for x in range (5000):
    weights = np.random.random(num_assets)
    weights /= np.sum(weights)
    pfolio_returns.append(np.sum(weights * returns.mean()) * 250)
    pfolio_volatilities.append(np.sqrt(np.dot(weights.T,np.dot(returns.cov() * 250, weights))))
    pfolio_weights.append(weights)
    
pfolio_returns = np.array(pfolio_returns)
pfolio_volatilities = np.array(pfolio_volatilities)
#pfolio_weights = pfolio_weights.tolist()

# pfolio_returns, pfolio_volatilities, pfolio_weights


portfolios = pd.DataFrame({'Return': pfolio_returns, 'Volatility': pfolio_volatilities})


portfolios.plot(x='Volatility', y='Return', kind='scatter', figsize=(10, 6));
plt.xlabel('Expected Volatility')
plt.ylabel('Expected Return')
plt.show()


max_return = max(pfolio_returns)
max_return

0.15181206184284307


max_rindex = np.where(pfolio_returns == max_return)
print(max_rindex)

(array([2754], dtype=int64),)


max_rvolatility = pfolio_volatilities[max_rindex[0][0]]
max_rvolatility

0.22375353322206148


max_rweights = pfolio_weights[max_rindex[0][0]]
max_rweights

array([0.01906369, 0.05031982, 0.00552441, 0.42967666, 0.49541542])


min_volatility = pfolio_volatilities.min()

min_volatility

0.2032442633026616


min_vindex = np.where(pfolio_volatilities == min_volatility)
min_vindex

(array([3964], dtype=int64),)


min_vreturns = pfolio_returns[min_vindex[0][0]]
min_vreturns

0.10898203038264713


min_vweights = pfolio_weights[min_vindex[0][0]]
min_vweights

array([0.20377067, 0.09930902, 0.18635316, 0.33062592, 0.17994123])


max_sr = []
max_sr =  [str(round(max_return * 100,3) ) + '%', str(round(max_rvolatility * 100,3)) + '%']
for i in range(num_assets):
    max_sr.append(str(round(max_rweights[i] * 100,3)) + '%')
max_sr

['15.181%', '22.375%', '1.906%', '5.032%', '0.552%', '42.968%', '49.542%']


min_vol = []
min_vol =  [str(round(min_vreturns*100,3)) + '%', str(round(min_volatility * 100,3)) + '%']
for i in range(num_assets):
    min_vol.append(str(round(min_vweights[i] *100,3)) + '%')
min_vol

['10.898%', '20.324%', '20.377%', '9.931%', '18.635%', '33.063%', '17.994%']


#Max Sharpe Ratio
MaxSharpeRatio = go.Scatter(
    name='Maximium Return',
    mode='markers',
    x=[max_rvolatility],
    y=[max_return],
    marker=dict(color='red',size=14,line=dict(width=3, color='black'))
)

#Min Vol
MinVol = go.Scatter(
    name='Mininium Risk',
    mode='markers',
    x=[min_volatility],
    y=[min_vreturns],
    marker=dict(color='green',size=14,line=dict(width=3, color='black'))
)

#Random portfolio
EF_curve = go.Scatter(
    name='Random Portfolios',
    mode='markers',
    x= pfolio_volatilities,
    y= pfolio_returns,
    marker=dict(color='blue',size=5)
)

data = [MaxSharpeRatio, MinVol, EF_curve]

layout = go.Layout(
    title = 'Portfolio Optimisation with the MPT',
    yaxis = dict(title='Annualised Return'),
    xaxis = dict(title='Annualised Volatility'),
    showlegend = True,
    legend = dict(
        x = .75, y = 0, traceorder='normal',
        bgcolor='#E2E2E2',
        bordercolor='black',
        borderwidth=2),
    width=800,
    height=600)

fig = go.Figure(data=data, layout=layout)
fig.show()


col = []
col =  ['Returns', 'Volatility']
for i in range(num_assets):
    col.append(pf_data.columns[i])


result_table = pd.DataFrame(columns  = [col], index=['maximum Return', 'minimun risk'])
result_table.iloc[0] = max_sr
result_table.iloc[1] = min_vol


result_table


sn.heatmap(corr_matrix, annot=True)
plt.show()


trace = go.Heatmap(
   x = assets,
   y = assets,
   z = corr_matrix,
   type = 'heatmap',
   colorscale = 'Viridis',
   hovertemplate = "%{z}"
)
layout = go.Layout(
    title = 'Correlation',
    yaxis = dict(title='Stocks'),
    xaxis = dict(title='Stocks'),
    showlegend = False,
    width=500,
    height=500)
data = [trace]
fig = go.Figure(data = data, layout = layout)
fig.show()

	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS	NCC.NS	ONGC.NS	RELIANCE.NS	SBIN.NS	WIPRO.NS
Date
2018-04-02	177.256470	1715.386230	3040.504150	211.317444	1219.384155	109.640846	132.522751	865.632080	240.150375	212.622787
2018-04-03	178.182709	1703.860352	3039.628174	211.522034	1214.318481	116.553490	130.718475	872.030151	244.394363	208.401917
2018-04-04	175.372116	1682.955444	3058.141113	211.603867	1183.056885	115.042786	129.171951	867.522522	241.272354	206.713547
2018-04-05	177.448090	1702.693848	3154.043213	212.463043	1212.858154	118.567772	130.681671	880.415527	252.979858	210.126953
2018-04-06	176.170578	1706.940063	3151.791504	213.444962	1201.676880	119.529121	130.313446	883.130005	253.370132	208.438614
...	...	...	...	...	...	...	...	...	...	...
2023-03-23	213.350006	2569.850098	2362.500000	380.649994	2206.649902	105.949997	152.800003	2247.800049	512.750000	361.500000
2023-03-24	208.449997	2559.899902	2313.350098	378.700012	2166.149902	101.900002	149.399994	2203.300049	505.350006	361.750000
2023-03-27	208.399994	2561.899902	2308.600098	380.200012	2154.000000	101.449997	150.250000	2237.550049	510.149994	361.250000
2023-03-28	208.899994	2580.149902	2249.500000	379.750000	2133.600098	102.800003	147.199997	2248.000000	508.100006	356.200012
2023-03-29	212.949997	2603.050049	2304.149902	382.750000	2147.449951	106.849998	149.350006	2234.699951	516.500000	359.049988

	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS	NCC.NS	ONGC.NS	RELIANCE.NS	SBIN.NS	WIPRO.NS
Date
2018-04-02	177.256470	1715.386230	3040.504150	211.317444	1219.384155	109.640846	132.522751	865.632080	240.150375	212.622787
2018-04-03	178.182709	1703.860352	3039.628174	211.522034	1214.318481	116.553490	130.718475	872.030151	244.394363	208.401917
2018-04-04	175.372116	1682.955444	3058.141113	211.603867	1183.056885	115.042786	129.171951	867.522522	241.272354	206.713547
2018-04-05	177.448090	1702.693848	3154.043213	212.463043	1212.858154	118.567772	130.681671	880.415527	252.979858	210.126953
2018-04-06	176.170578	1706.940063	3151.791504	213.444962	1201.676880	119.529121	130.313446	883.130005	253.370132	208.438614
...	...	...	...	...	...	...	...	...	...	...
2023-03-23	213.350006	2569.850098	2362.500000	380.649994	2206.649902	105.949997	152.800003	2247.800049	512.750000	361.500000
2023-03-24	208.449997	2559.899902	2313.350098	378.700012	2166.149902	101.900002	149.399994	2203.300049	505.350006	361.750000
2023-03-27	208.399994	2561.899902	2308.600098	380.200012	2154.000000	101.449997	150.250000	2237.550049	510.149994	361.250000
2023-03-28	208.899994	2580.149902	2249.500000	379.750000	2133.600098	102.800003	147.199997	2248.000000	508.100006	356.200012
2023-03-29	212.949997	2603.050049	2304.149902	382.750000	2147.449951	106.849998	149.350006	2234.699951	516.500000	359.049988

	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS	NCC.NS	ONGC.NS	RELIANCE.NS	SBIN.NS	WIPRO.NS
Date
2018-04-02	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2018-04-03	0.005225	-0.006719	-0.000288	0.000968	-0.004154	0.063048	-0.013615	0.007391	0.017672	-0.019851
2018-04-04	-0.015774	-0.012269	0.006091	0.000387	-0.025744	-0.012961	-0.011831	-0.005169	-0.012774	-0.008102
2018-04-05	0.011838	0.011728	0.031360	0.004060	0.025190	0.030641	0.011688	0.014862	0.048524	0.016513
2018-04-06	-0.007199	0.002494	-0.000714	0.004622	-0.009219	0.008108	-0.002818	0.003083	0.001543	-0.008035
...	...	...	...	...	...	...	...	...	...	...
2023-03-23	-0.002804	-0.005553	0.003952	0.007677	-0.003297	0.010009	0.008248	-0.012650	-0.016873	-0.013373
2023-03-24	-0.022967	-0.003872	-0.020804	-0.005123	-0.018354	-0.038226	-0.022251	-0.019797	-0.014432	0.000692
2023-03-27	-0.000240	0.000781	-0.002053	0.003961	-0.005609	-0.004416	0.005689	0.015545	0.009498	-0.001382
2023-03-28	0.002399	0.007124	-0.025600	-0.001184	-0.009471	0.013307	-0.020300	0.004670	-0.004018	-0.013979
2023-03-29	0.019387	0.008876	0.024294	0.007900	0.006491	0.039397	0.014606	-0.005916	0.016532	0.008001

	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS
Date
2018-04-02	177.256470	1715.386230	3040.504150	211.317444	1219.384155
2018-04-03	178.182709	1703.860352	3039.628174	211.522034	1214.318481
2018-04-04	175.372116	1682.955444	3058.141113	211.603867	1183.056885
2018-04-05	177.448090	1702.693848	3154.043213	212.463043	1212.858154
2018-04-06	176.170578	1706.940063	3151.791504	213.444962	1201.676880
...	...	...	...	...	...
2023-03-23	213.350006	2569.850098	2362.500000	380.649994	2206.649902
2023-03-24	208.449997	2559.899902	2313.350098	378.700012	2166.149902
2023-03-27	208.399994	2561.899902	2308.600098	380.200012	2154.000000
2023-03-28	208.899994	2580.149902	2249.500000	379.750000	2133.600098
2023-03-29	212.949997	2603.050049	2304.149902	382.750000	2147.449951

	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS
Date
2018-04-02	NaN	NaN	NaN	NaN	NaN
2018-04-03	0.005225	-0.006719	-0.000288	0.000968	-0.004154
2018-04-04	-0.015774	-0.012269	0.006091	0.000387	-0.025744
2018-04-05	0.011838	0.011728	0.031360	0.004060	0.025190
2018-04-06	-0.007199	0.002494	-0.000714	0.004622	-0.009219
...	...	...	...	...	...
2023-03-23	-0.002804	-0.005553	0.003952	0.007677	-0.003297
2023-03-24	-0.022967	-0.003872	-0.020804	-0.005123	-0.018354
2023-03-27	-0.000240	0.000781	-0.002053	0.003961	-0.005609
2023-03-28	0.002399	0.007124	-0.025600	-0.001184	-0.009471
2023-03-29	0.019387	0.008876	0.024294	0.007900	0.006491

Project - Stock Portfolio Optimization¶

Course : Financial Analytics¶

Group : 3¶

Importing Data¶

Prediction With LSTM:¶

Sending Data for Prediction:¶

Picking best performing stock¶

Returns¶

Random Weights¶

Mean Annual Returns¶

Annual Variance¶

Annual Covariance¶

Correlation¶

Markovitz Portfolio Theory¶

5000 Random weights¶

Max Sharpe ratio and Min volatility table with required weights of stocks¶

Correlation Graph¶

	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS
COALINDIA.NS	1.000000	0.231154	0.274088	0.300004	0.344331
HDFC.NS	0.231154	1.000000	0.380305	0.381712	0.546168
HEROMOTOCO.NS	0.274088	0.380305	1.000000	0.278446	0.402051
ITC.NS	0.300004	0.381712	0.278446	1.000000	0.372967
LT.NS	0.344331	0.546168	0.402051	0.372967	1.000000

	Returns	Volatility	COALINDIA.NS	HDFC.NS	HEROMOTOCO.NS	ITC.NS	LT.NS
maximum Return	15.181%	22.375%	1.906%	5.032%	0.552%	42.968%	49.542%
minimun risk	10.898%	20.324%	20.377%	9.931%	18.635%	33.063%	17.994%